← The American Hiring System

Sources & Methodology

Not every statistic in this article deserves equal trust. This is the reference node — which numbers are load-bearing, which are directional-only, and which claims were deliberately excluded because the research behind them doesn't hold up.

Published August 1, 2026·Last revised August 1, 2026

This node sits outside the main circle on purpose. It isn’t a section of the argument — it’s the record of how confident this article is in each part of that argument, and the full source list behind every claim made in the six sections around it.

What It Is

A tiered confidence rating for every category of source used in this topic, plus the specific claims that were deliberately excluded or corrected rather than repeated at face value.

How It Works

TierSource typeHow to read it
1Statistical agency / administrative dataBLS (JOLTS, Employment Situation, union membership, disability labor force data), EEOC, NY Fed and St. Louis Fed research notes, Stanford Digital Economy Lab’s ADP payroll microdata. Cited directly, highest confidence.
2Research orgs / think tanks with disclosed methodologyNFIB, EPI, Burning Glass Institute + Harvard Business School, Institute for Justice, NELP, IZA and Nature Human Behaviour academic papers, AARP, Staffing Industry Analysts, NCEE/Brookings/Urban Institute apprenticeship research. Cited with attribution; trend direction weighted more heavily than precise point-in-time levels for self-selected-sample surveys like NFIB.
3Corporate/vendor research and industry pressZipRecruiter and Indeed benchmark data, SHRM cost-per-hire figures, recruiting-industry burnout surveys, time-to-hire aggregators, resume-fraud self-report surveys. Directional only, flagged explicitly wherever used — ranges preferred over point estimates.
4Explicitly weak or discardedSmall self-report surveys with likely leading questions, and one specific mathematical error debunked directly below.

Ratings apply to source categories, not individual claims — see specific flags below for exceptions.

Two source types get elevated above their tier by design. Correspondence and audit-study findings — the age-discrimination callback study (29%/47% fewer callbacks by identical resume, varied age) and the caregiving-gap study — are controlled experimental designs, not surveys, and are treated throughout this topic as stronger evidence than any self-report figure they sit next to. Distinguishing stated reasons from measured causes is a discipline applied everywhere in this topic, most visibly to Challenger’s AI-layoff attribution data (an employer’s chosen public explanation, not a controlled measurement of what actually caused a layoff).

Claims deliberately excluded or corrected rather than repeated:

  • “93% of HR professionals post ghost jobs” (LiveCareer survey). Small sample (918), self-reported, likely leading question design. Not used as a lead statistic anywhere in this topic. Greenhouse’s platform-derived 18–22% figure is used instead — see Macro Labor Data.
  • “Openings minus hires equals ghost jobs.” A stock/flow category error, not a real finding — openings is a stock measured on the last business day of a month, hires is a flow summed across the whole month, and the gap between them exists in every JOLTS release ever published. This argument is actively corrected, not repeated, in Macro Labor Data.
  • “250 applications per job posting” / “applicants doubled from 46 to 95 since 2021.” Circulates widely across recruiting-vendor blogs with no traceable primary source or disclosed sample frame. Not cited anywhere in this topic as a point estimate.
  • Brookings’ “44% of workers earn low wages” figure. Dates to November 2019, pre-pandemic wage compression at the bottom of the distribution — used in Wage & Job Quality only as historical/definitional baseline, never as a current-year estimate. The Job Quality Index, current through mid-2026, carries the present-day trend claim instead.

A third caveat applies specifically to the Solutions section. Growth and adoption statistics for the fractional-executive market (percentage of small businesses using fractional leadership, market-size projections) circulate widely but come almost entirely from vendor and marketing content with undisclosed methodology — directionally consistent with a real, fast-growing market, but not precise enough to state as point estimates. The Fractional Talent Exchange and Talent-Sharing Cooperative solutions cite this category of source as Tier 3 explicitly, the same way the rest of this topic treats recruiting-vendor data.

Two standing caveats apply across every section. JOLTS runs roughly a month behind real time — May 2026 was the most recent release at the time of research, and later data may shift specific figures without changing the overall pattern described. And NFIB, SHRM, and most recruiting-vendor data reflect self-selected or member populations, which is useful for trend direction and risky for precise point-in-time levels stated as if population-representative.

Why It Matters

Treating all statistics as equally solid is how a well-researched article quietly becomes a misleading one. The single most consequential methodological choice in this topic is elevating two controlled studies — on age discrimination and on caregiving-gap penalties — above the flashier but weaker survey statistics that surround them, and being explicit, in the sections themselves, about which numbers are ranges rather than points.

DimensionStatusNotes
Government/Admin DataHigh ConfidenceBLS, EEOC, Federal Reserve research, and ADP administrative payroll microdata — cited directly throughout.
Vendor/Survey DataDirectional OnlySHRM, staffing-industry, and recruiting-vendor figures are used for trend direction, flagged explicitly, and given as ranges rather than point estimates.
Debunked ClaimsActively CorrectedThe ghost-jobs stock/flow argument and the unsourced '250 applications per posting' figure are named and corrected rather than repeated.

Full Bibliography

Government / statistical agencies

Research institutions / think tanks

News / industry press

Vendor / secondary aggregator sources (Tier 3 — directional only)

  • Recruiting-industry burnout/workload sources: Joveo, Phenom, Recruiterflow, Ninjahire (2026)
  • SHRM-citing cost-of-hire aggregator sites, cross-referenced for consistency
  • Time-to-hire benchmarking aggregators (multiple 2026 sources, wide variance noted)
  • LiveCareer HR professional ghost-jobs survey (Tier 4 — discarded as a lead statistic; see above)
  • Fractional-executive-market growth and adoption statistics (multiple 2026 vendor/marketing sources) — directional only; not cited as point estimates in the Fractional Talent Exchange solution
  • Contractor-vs-W-2 rate-premium calculators and guides (multiple sources, 2026) — cross-referenced for consistency on the general 1.3–2x break-even multiple; treated as standard methodology rather than a single citable study, and explicitly scoped in the Fractional Talent Exchange solution to commoditized hourly contracting rather than value/project-priced expert work
  • Billable-hour efficiency-penalty commentary (consulting and legal-industry sources, 2026) — vendor/industry-blog sourced, Tier 3, but the core critique appears independently and repeatedly across both consulting and legal-billing literature; cited in the Fractional Talent Exchange solution
  • 2025 consultant pricing-model survey (project-rate/value-based/hourly split) and an older academic literature review of B2B value-based-pricing adoption (1983–2006 studies) — cited together in the Fractional Talent Exchange solution specifically to avoid overstating how many professionals use pure value-based pricing

Revision History

DateChanges
August 1, 2026Add Pelotech, Fraction, and IT staffing-agency sourcing for the Talent-Sharing Cooperative solution
August 1, 2026First published

Discussion